Paper prepared for 20th EGOS Colloquium July 1-3, 2004

The Organization as a Set of Dynamic Relationships

Ljubljana University, Slovenia

Sub-theme 18: Corporate Social Responsibility, Corporate Citizenship and Politics of

Stakeholder Influence

Corporate Social Responsibility: The Problem of Securing External

Legitimacy in a Globalized World

Claes Ohlsson and Stefan Tengblad

Gothenburg Research Institute

School of Economics and Commercial Law. Göteborg University

P. O. Box 600

SE-405 30 Gothenburg, Sweden

Contact person: Stefan Tengblad

E-mail: stefan.tengblad@gri.gu.se

Ph. +46 31 773 56 12, Fax +46 31 773 56 19

1

Corporate Social Responsibility: The Problem of Securing External Legitimacy in a

Globalized World

Claes Ohlsson and Stefan Tengblad

Gothenburg Research Institute

This paper presents the results from a textual analysis of "Letters from the CEOs" in leading Swedish

corporations. The sample consists of letters from annual reports for the years of 1981, 1991 and 2001.

The purpose of the paper is to discuss how the textual representations of issues related to corporate

social responsibility (CSR) have changed over the period. The results show, not surprisingly, a

substantial increase on a number of topics that can be linked to the general CSR-discourse in the 2001

sample. But the rise of a CSR-discourse is related to a drop of another discourse related to social

responsibility, in which the CEOs used to comment upon the social, economic and political

development of their native country. It is therefore doubtful to claim that business leaders of today feel

a larger social responsibility than in the past. In the end section of this paper we discuss the

consequences of the shift of responsibilities from a narrower national arena towards a globally

dispersed community of stakeholders. It is concluded that the CSR discourse, firmly rooted in a self-

regulation philosophy, have not yet emerged as an influential movement in Sweden that affects the

overall business conduct in a more substantial manner. The CSR-philosophy seems so far to be a more

liberating than restraining factor for those Swedish companies who have moved their operations across

the national borders.

Introduction

Corporate Social Responsibility (CSR) is a topic that have gained a widespread attention over

the last 5-10 years. The number of hits on the concept on Internet, using the search engine

www.google.com, exceeds half a million (May 2004) and a large number of companies have

established special departments for dealing with issues related to CSR. CSR adheres to a self-

regulation philosophy aiming to achieve capitalism with a human face, by the means of

ethical awareness and a readiness to contribute to societal development by voluntarily and

2

pro-active behaviour. CSR is thus aimed to counter-weight potential drawbacks of free-

market capitalism such as excessive exploitation and mistreatment of workers, the

environment and local societies.

Political and organizational initiatives have been instrumental for the construction and

diffusion of the CSR-discourse. In 1999 the Global Reporting Initiative (GRI) presented a set

of general guidelines for measuring economical, political and social performance for all kinds

of business, governmental and non-governmental organizations. The aim was to stimulate

different kinds of organizations to report upon and to take action to improve the social effects

of its operations. In the following year, several influential guidelines of CSR were also

presented; the UN's Global Compact, the ILO's Tripartite Declaration of Principles

concerning Multinational Enterprises and Social Policy and the OECD Guidelines for

Multinational Enterprises. Equally important in the setting are the EU guidelines for

Corporate Social Responsibility (the Green Paper), which were adopted by the European

Union in 2001.

These initiatives for launching more encompassing social responsibility than the legal

requirements have also affected the business and organizational life of Sweden. One effect

was that business ethics and social responsibility issues became frequently discussed in media

in this period (1999-2001). One example is a business press article from November 2000, in

which the respected spokesman in business affairs Björn Svedberg argued that Swedish

companies should resume a global responsibility for environmental and human rights issues.1

The former CEO of the engineering giant ABB, Göran Lindahl, CSR also received much

attention for his assignment as a special advisor in CSR to the UN Secretary-General, Kofi

Annan.

The large attention given to CSR-related issues can also be illustrated by the fact that when

the CEOs were authoring their CEO letters for the annual reports of 2001, the Swedish

government launched a national CSR initiative called the Swedish Partnership for Global

1 Mr Svedberg is a former CEO of Ericsson and a board member in several large companies at this time. Source:

Svenska Dagbladet, 21 November, 2000.

3

responsibility in big advertisements. Swedish companies were in theses advertisements

invited to adhere to the OECD guidelines and the UN Global Compact.2

In this paper we will examine the impact of the CSR discourse in CEO letters. We will also

examine whether the CSR discourse provides a new discursive space for the CEOs to behave

in a more socially responsible way, or if the rise of the CSR discourse replaces other topics,

which are related to stakeholder responsibility. Our findings indicate that the CSR-discourse

have replaced a societal discourse related to the national arena. There is also a remarkable

shift in geographical references in the examined CEO letters from words such as Sweden,

abroad and export/import to words as USA, Asia, and global.

In order to investigate how textual representations of topics related to CSR have changed we

have analysed "The letters from the CEO's" in annual reports of 15 large Swedish

corporations for the years of 1981, 1991 and 2001. The companies analyzed primarily consist

of leading Swedish industrial, service and financial companies such as Electrolux, Ericsson,

Handelsbanken, Hennes & Mauritz, Skanska, SKF, SEB and Volvo.

Data selection and methodology

Annual reports are produced not least in order to create confidence and interests about the

company. Old shareholders are to be convinced to keep their shares and prospective

shareholders should be inclined to make an investment. To successfully make use of the

annual report to attract shareholders, we argue, the messages communicated need to respond

to the values of the shareholders. Moreover, annual reports are also perhaps the most

important document for companies to articulate important messages to a wider body of

stakeholders, such as employees, government agencies, interest groups, etc. In order to gain

approval and legitimacy from such groups, articulated messages in annual reports are often

formulated strategically in order to respond to values of these groups. Thus what is written in

the annual reports can be anticipated to reflect normative values about what constitutes good

governance among the general public of shareholders/stakeholders. If for instance the general

2 Dagens Industri, p 3, March 7, 2002.

4

public, and the majority of shareholders, expect the companies to devote a lot of energy in the

field of CSR, the CEO letters are expected to contain information about CSR.

The CEO letters are not mainly the very personal thoughts of chief executives, they belong to

a genre3 were the writers write in order to meet the readers expectations of what modern and

competent chief executives should be concerned about. The change of annual reports as text

type and their development as genre during the 20 year-scope is in this paper seen as a result

of changes in the corporate discourse practice. An eclectic range of studies concerning annual

reports and CEO letters can be compared with our study, which includes Hyland (1998) who

focuses on meta-discourse of CEOs and how a positive message is manifested in the annual

report in order to influence the reader. Also, Thomas (1997) studies the linguistic structures of

management messages in annual reports and how positive and negative messages are

communicated differently. Finally, Yuthas et al (2002) is an attempt to investigate ethical

characteristics of discourse found in annual reports.

There are no legal requirements in the Swedish Company Act that a special letter from the

CEO (or 'the Chairman's report') should be included. But nevertheless there are strong

expectations that an annual report in a major company should start with a letter from the CEO

(or board of directors) (Skulstad 2002). Its theme is always the affairs of the company of the

previous year from the author's/the CEO's view and often also an outlook over the near future.

Their structure and form are usually quite similar. A typical CEO letter contains somewhere

between 1.000 to 1.500 words and is usually illustrated with a picture of the CEO. The mass

of texts from the 41 analysed CEO letters contains 48 700 words. The CEO letters are in

general longer in 2001 than in the other two periods and the amount of text from 2001 is

subsequently larger. The differences in size between the texts from the three periods must be

held in mind when analysing them. The details, length and size of the selected CEO letters are

presented in Appendix B.

The main condition for selecting the companies is that they have been listed at the Stockholm

Exchange throughout the entire period. We have also strived to include the largest of the

listed companies measured by numbers of employees and by capital assets, because such

companies receive large media attention and are more scrutinized by the general public than

3 Our use of the notion of genre follows Swales (1990).

5

smaller companies. Therefore, the well-known company can be expected to adhere to the

external expectations to a larger extent than for instance very small companies on the

Exchange market. A general description of the selected companies is provided in Appendix A.

The selection of years are firstly made in order that each of three decades are represented in a

10-year interval, and secondly that the economic conditions are about the same. The CEO

letters can otherwise be affected by business cycles and the CEOs may be occupied by

different problems in boom and bust times. 1981, 1991 and 2001 were all down-turn years.

1981 was a year plagued by the effects of the second oil-crises and appeared before the large

Swedish devaluation took place in the autumn of 1982. 1991 was a very problematic year for

the Swedish economy and the unemployment reached for Swedish conditions very high

levels. Finally, 2001 was characterized by weakened demand, severe problems in the IT

sector and political instability in the wake of the horrifying terror attacks on September 11.

Our selection of annual reports from over two decades draws attention to the change and

growth of the genre of annual reports. In 1991 and 2001 all the selected annual reports contain

CEO letters but in the 1981 sample, only 11 of 15 annual reports contain a CEO letter. This

can been seen as a part in the development towards more elaborated annual reports in Sweden

and elsewhere. Our sample of texts shows that annual reports gets longer over the period and

also contains more kinds of text beside the legally stipulated information and possible CEO

letter. In the 2001 sample, several companies publish two separate parts of the annual report,

one part containing the financial information and the other presenting other information in

narrative sections. In line with longer annual reports and the introduction of new types of

information, the reports have become more expensive-looking and are in 2001 printed in a

style that resembles glossy magazines.

Discourse and discourse analysis

This paper uses a linguistic framework of discourse analysis with an emphasis on the lexical

level of language. The use of discourse analysis is widespread in many disciplines and it is

usually difficult to distinguish a generally accepted notion of method when it comes to the

field of discourse studies. The very concept of discourse is also viewed upon in several ways.

Grimshaw (2003: 27) sees two very different views or approaches on the notion of discourse.

6

The first approach consists in seeing the discourse as "[...] an ideological 'bundle', a

subculture, or even an arena of special interaction." (ibid.). The second approach treat or see

discourse as "[...] spoken or written text in a language, intended for use in the accomplishment

of social ends of users'" (ibid.). The study of discourse, according to this second approach, is

also the study of language in use.

In the field of linguistics, discourse is generally seen as spoken or written language – what can

be called talk and text – and the focus of discourse studies is on language in use, as of

Grimshaw (opt. cit). The traditional linguistic way of analysing discourse can be described as

descriptive in contrast to the critical approach of discourse analysis. The latter is generally

associated with the work of Norman Fairclough (1992, 1995) and the CDA-school (Critical

Discourse Analysis), where power and status are key-notions. In descriptive discourse studies,

"background" and "knowledge" tend to serve as general categories of explanation for

language phenomena and they are usually not closely scrutinized or problematized.

With the critical approach of this paper, the language use of CEOs is seen in close connection

to the setting, context and (historical) background. In the critical analysis of discourse, it is

emphasized that the discourse itself reproduces what Fairclough calls macro structures

(1995:43). Fairclough adopts a three-dimensional conception of discourse and as method of

discourse analysis. The three dimensions are (i) text, (ii) discourse practice (text-production

and text-interpretation) and (iii) socio-cultural practice (ibid: 97). In this case, both the

production and need of annual reports as well as the development of CSR can be seen as

socio-cultural practices. The annual report and the letter from the CEO is the discourse

practice, which of course includes the text itself (the actual discourse). Although, within the

theoretical frame of CDA, our focus is mainly descriptive-empirical at this level.

Text and corpus analysis

Beside being inspired by the CDA-school of linguistics, this paper attempts to combine CDA

as described above with the strength of corpus analysis on lexical (or word-) level. A corpus

can be described as a body of text or texts, which can be linguistically analysed in several

ways (see Stubbs 1996). For our purpose, the selected annual reports from the three time-

periods where quantified with the help of computer software (Concorder Pro). This

quantification gives us the possibility to sort all words in the texts by either frequency or in

alphabetical order, which enables a description of each text or every text from one of the

7

selected periods. The quantification also gives us a tool to compare the texts from 1981, 1991

and 2001 to each other. The software can produce an index of all words and also a

concordance of selected texts. A concordance is a presentation of a selected word in its

context where the word searched for is centred, which makes it easier to quickly analyse a

larger body of text than a method of close-reading alone.

Our method for selecting words has partly been initiated by 'buzz words' that can be found in

the CSR-paradigm (such as documents from institutions like the European Commission or the

UN but also from companies and media), and partly from the lists of word frequency and

alphabetical word order derived from the material (see Piper 2000 for a similar methodology).

The production of concordances has led us to tables of salient words and clusters of words,

which enables us both to see how a word pattern changes over time but also how a word like

"customer" is used in different compound words and similar constructions. This sorting of

words also helps us to find our way back into the text and closely analyze samples of the text

in the annual reports.

Empirical findings

The empirical findings of our study and their presentation are grouped in three themes as of

below.

1. The emergence of the CSR discourse.

2. The decline of a societal discourse related to the national arena.

3. The shift in geographical perspective.

Every presentation includes a table of salient words and word clusters. A further explanation

of the single and compound words that make up the presented clusters follows the tables.

Beside the discussion of compound words, we present salient examples from the texts. The

examples we present are all in English, either the company's translation from an English

version of the annual report or our translation when an English version not is available.

8

The emergence of the CSR discourse

When scrutinizing the word frequency list generated from the selected CEO letters, no

immediate CSR-related words can be found among the most used words. However, it is quite

clear, as Table 1 shows, that the CSR discourse has its breakthrough in the 2001 sample with

nearly 50 occurrences of words denoting the companies' social responsibility as of the

emerging CSR international trend.

Table 1. Word clusters denoting CSR-discourse

Word cluster occurrences 1981 occurrences 1991 occurrences 2001

environment/environmental 0 13 8

ethics/ethical 0 0 3

long-term (orientation) 0 0 12

responsible/responsibility 7 2 10

stakeholder(s) 2 0 2

sustainable/sustainability 0 0 9

value (other than financial) 0 0 4

Total 9 15 48

An important part of our analysis are the compound constructions of words (mostly nouns)

denoting a discourse practice like CSR. In the case of CSR, most salient words do not occur in

several compound constructions. The main exception is the cluster of "environment" where

some ten different compound words can be distinguished.4

It seems clear that there are certain words that denote a CSR discourse and that those words

mostly are found in the 2001 sample of CEO letters. The exceptions, which stand out in Table

1, are the word clusters of "responsible/responsibility", "stakeholder" and

"environment/environmental". The most common word in the "responsibility" cluster of 1981

is "owner responsibility" with four occurrences, a word that not appears in the 1991 or 2001

samples. However, a shift from outspoken owner responsibility to a more general notion of

responsibility can be traced in the word cluster of "responsibility" from 1981 to 2001. Also

4 Among these are miljöarbete (environmental work), miljöfrågor (environmental issues) and miljöpåverkan

(environmental impact). In the cluster of "responsibility", interesting compounds are helhetsansvar (complete or

total responsibility) and kundansvar (customer responsibility).

9

interesting are the occurrences of word forms about "environment" where a peak can be seen

for the 1991-sample of CEO letters. It is possible to ascribe the CEO letter focus on

environmental issues in 1991 to a general environmental movement in Swedish society of the

early 1990's. During these years, the Green Party reached the parliament for the first time

(1988) and several incidents occurred where the environment in Sweden was damaged or

threatened.5

We will provide a number of illustrative quotations in order to provide an understanding of

the content of the CSR discourse that can be found in the 2001 CEO letters. Example 1 below

shows an example from the 'ethics-cluster' where the word ethical appears together with the

word responsible.

(1) We owe it to our employees, our clients and our shareholders - and especially the public - to

operate this business in a responsible and ethical way. That is what we try to do - every day,

everywhere. (Our italics, original English version, Skanska 2001)

The emphasis on being a responsible company that operates in an ethical way is not supported

by specific examples on actions taken. Another example (2) that clearly illustrates the

emergence of a CSR discourse in the CEO letters comes from the manufacturing company

Trelleborg.

(2) We want to create and preserve a durable business concept that is based on our core values. We

associate ourselves with the concept of sustainable growth. We are working with these issues on a

continuing basis, and during the year we also formulated a Code of Conduct, with certain rules

showing how we can ensure sustainable growth with our various stakeholders. (Our italics,

original English version, Trelleborg 2001)

The importance given by Trelleborg to the concept of core values is related to the creation of

sustainable growth and the formulation of a Code of Conduct. Trelleborg, and the insurance

company Skandia also exemplify an increasing use of the word value(s) in non-financial

settings (example 3 and 4).

5 One such incident was the mass death of seals on the Swedish west coast, which became an issue in the election

campaign of 1988.

10

(3) In a fast-growing industrial groups like Trelleborg, with both cultural differences and different

values, a company's common values assume special importance. (Our italics, original English

version, Trelleborg 2001)

(4) It is therefore important that we do everything in our powers - as individuals as well as a

company - to restore values such as confidence, security, and belief in the future. (Our italics,

original English version, Skandia 2001)

To conclude, the emergence of a CSR discourse is evident in the 2001 sample of CEO letters

when compared with the 1981 and 1991 texts. Not least is it interesting to see the increasing

use of the world of long-term at a period characterized by rapidly increasing short-term

pressures to provide superior financial returns (Tengblad 2003). It is also worth noting that the

selected CSR-related words only represents about 0.3 percent of the total text.

The decline of a societal discourse related to the national arena

Before the advent of the quite general and international CSR discourse about ethics,

responsibility, sustainability, etc, it was not unusual for CEOs to include comments about the

political and economical development of Sweden in their letters. These comments, which can

be viewed as a part of a larger societal and political discourse, contained remarks about how

the CEOs were viewing the national development, such as what should be done in order to

break the deteriorating economic situation of Sweden at the early eighties and nineties. This

discourse included a concern about employment, balance of trade as well as the issue of

inflation. Words denoting the Swedish state and government were also used within the

discourse along with the theme of politics in general.

11

Table 2. Word clusters denoting a societal/national discourse

Word cluster occurrences 1981 occurrences 1991 occurrences 2001

(national) competitiveness 1 2 0

(un)employment 3 2 0

export/import 13 4 1

inflation 14 2 0

politics/politician 5 15 2

the Government 3 3 0

the National bank 5 0 0

the State 6 2 0

unions/co-determination 6 0 0

bargaining 4 0 0

Total 60 30 3

It is obvious that the societal discourse related to a national arena has ceased to be in focus in

the CEO letters over time. The drop of word occurrences is largest between the years of 1991

and 2001. In the 2001 sample the hits are so few that it makes no sense to talk about a

discourse anymore.

The pattern of decline is visible in all selected word clusters. A closer look at the occurrences

of the "export" and "import"-clusters, reveals that the subject of Swedish export/import was

frequently used in compound constructions 1981 but rarely in the later samples. Such 1981

compounds are export revenue, export market and export increase. The lesser common import

compounds are import restrictions and import pressure. The export compound words are in

general positive words while the two import compounds are portraying negative effects. The

"export/import"-theme can also be linked to the theme of "competitiveness", which focus on

the competitiveness of Sweden and Swedish industry. This is invoked in both the 1981 and

1991 samples but not at all in the 2001 CEO letters.

The "inflation" cluster is also mainly a part of the message in the 1981 CEO letters. The drop

is very clear with 14 occurrences in 1981, compared with two in 1991 and none in 2001. This

can be seen in relation to the previous reference of the inflation as a major issue in Swedish

politics during the 1980's and early 1990's, which also goes in line with the compound words

12

that can be found in the "inflation" word cluster. Such compound words translate to rate of

inflation, inflation development and fight of inflation.

The word clusters of "the state" and "politics/politician" show a similar decrease from 1981 to

2001 with one prominent exception of "politics/politician" for 1991. In the sample of 1991

CEO letter, this theme presents 15 occurrences. The relatively turbulent politics of the early

1990's stand out as a possible explanation to the increase of politically related word

occurrences in the 1991 CEO letters. When scrutinizing the 1991 sample, it stands clear that

the right-block political party win in the national election of 1991 has made an impact in the

CEO statements. The dramatic and, at the time, recent changes in the countries of the old

communist block are also commented within the theme of politics. The East European

countries are described as possible new markets for the companies and the political changes,

local and international, are depicted in a generally positive manner in the 1991 sample.

Not least interesting in the sample is that the CEOs do not mention anything about unions, co-

determination or bargaining expect in the 1981 sample where these words were quite

frequently mentioned. On the whole the 1981 references to politics, the State, the Government

and unions are in general concerned with the national level and the CEO's are stating their

disagreement with the current political situation and its effects on business and trade. There

are also references made to possible (positive) changes in jurisdiction made by the

government in 1981 and also to subsidized projects where the State is involved. The debate on

wage earner's investment funds was also highly infected during the 1980's in Sweden and

these union-based funds are criticized in the CEO letters (Example 5):

(5) The centrally controlled union wage earner's funds now debated, would in the long run be a

catastrophe for the industry of Sweden. (Our translation, Atlas Copco 1981)

The comparison with the 1981 and 1991 samples to the 2001 CEO letters is clear when it

comes to the word clusters of "politics/politician", "state" and "government". Virtually no

references are made to national politics in the 2001 sample. One occurrence in the "politics"-

cluster of 2001 also typically refers to global politics with the September 11 attacks as

background, which is seen in Example 6:

13

(6) In addition to weaker economies and heightened risks in the global political arena, capital

markets in recent months have been influenced by greater uncertainty about the integrity of

corporate accounting methods. (Our italics, original English version, Investor 2001)

The shift in geographical perspective

The decline of the national and societal discourse that emphasises national politics and natio-

nal stakeholders (the State, the unions) is evident in the CEO letters. The main reason for this,

we believe, is the ongoing globalization of economies, companies and even cultural artefacts.

As Table 3 (below) illustrates how the relative importance of Sweden as a home base has

decreased, and this holds true regardless if one consider sales, employment or ownership.

Table 3. Sales of manufacturing, construction and retail companies in the sample

Approx. percentage of total sales

in Sweden

Approx. percentage of total sales

outside Sweden

Company

1981 1991 2001 1981 1991 2001

Atlas Copco 9 6 3 91 94 97

Electrolux 27,5 11,5 na 72,5 88,5 na

Ericsson 21 13 3 79 87 97

H&M 75 47 22 25 53 88

Sandvik 7 6 5 93 94 95

SCA 31 18 8 69 82 92

Skanska 78 73 17 22 27 83

SKF 7 4 4 93 96 96

Trelleborg 50 51 8,7 50 49 91,3

Volvo 23 14 na 77 86 na

The shift of importance in favour of an international market is clear for the manufacturing,

construction and retail companies in our study. In 1981, the clothing company H&M had 75%

of their total sales share in Sweden. In 2001, the total sale percentage in Sweden has

decreased to 22% for H&M. Still, this is the largest national sales percentage for any of the

selected companies in our study. In some companies, such as Volvo and Electrolux, the

Swedish sales figures of 2001 are not seen important to report separately, but are included in

Northern or Western Europe.

14

The shift from Sweden to an international presence for these companies is also revealed when

looking at the numbers of employees in and outside Sweden for the selected sample years

(Figure 1 below).

Figure 1. Number of employees in the selected 15 companies, 1981, 1991 and 2001

The 15 companies have all had a significant presence outside Sweden since several years but

a clear trend can be seen in the number of employees both in and outside Sweden. The first

amount has decreased from over 200.000 persons in 1981 to below 150.000 in 2001. The

number of employees outside Sweden of the 15 companies has, on the other hand, increased

from some quarter of a million in 1981 to well over 400.000 in 2001. The increasing

importance of an international presence can further be supported by looking at ownership of

shares in Swedish listed companies. In 1983, 8 percent of the shares in listed companies were

owned by non-Swedish owners, in 1991 the figure had risen to 12 percent and in 2001, 35

percent of the shares at the Stockholm Exchange were owned outside Sweden.6

6 This information comes from Statistics Sweden (www.scb.se). The earliest report year is 1983.

0

50000

100000

150000

200000

250000

300000

350000

400000

450000

Employees in Sweden Employees outside Sweden

1981

1991

2001

15

Our third, and final theme of lexical evidence from the CEO letters concerns the content of

geographical references. Two subgroups of word clusters are distinguishable. The first

concerns geographical places such as names of nations and continents or areas of the world.

The other subgroup comprises geographical entities of a more abstract nature. The

representation of words in the sampled years of CEO letters is shown in Table 4 below.

Table 4. Word clusters of geographic places and entities

Word cluster occurrences 1981 occurrences 1991 occurrences 2001

Sweden/Swedish 83 85 28

Nordic/Scandinavian 1 45 28

Europe/European 32 70 50

North America/USA 23 37 77

Asia/Japan 5 14 15

abroad 10 3 1

foreign 13 7 3

international(ly) 35 34 6

global(ly) 1 5 34

the World 31 38 56

The geographical references made in the CEO letters show a remarkable shift from a national

perspective to a regional perspective (Scandinavia/Europe), and particularly in the 2001

sample, to a global perspective. The declined referring to Sweden can be illustrated by the

fact that Sweden's share of the selected geographical categories decreased from 57% in 1981,

to 34% in 1991, and to a meagre 14% in the 2001 sample. References made to Scandinavia

and Europe peaks in the 1991 sample while references to North America/USA dominate in

the 2001 sample.

To some extent this pattern reflects extraordinary events. The September 11 events accounts

for some 15 references in the 2001 CEO letters. Thus, the whole increase of the USA-cluster

cannot be attributed to the terror attacks and it is clear that the references to USA in general

has increased for every sampled period. This indicates a shift from the national scene towards

16

Europe and more prominently USA/North America. Sweden's advances to join the European

Community (EC) account for the increased word occurrences in the "Europe/European"-

cluster of 1991. There are 20 occurrences of compound words denoting the EC in the 1991

sample. The dramatic political re-structuring of Eastern Europe is also part of this increase for

1991 as 10 references are made to Eastern Europe.

The "Asia/Japan"-cluster increases too from 1981 to 1991 and 2001. A shift can be noticed

between 1991 and 2001 where Asia in general becomes the commonest word in 2001,

compared with Japan for the 1991 sample.

The findings in the second subgroup underline the shift in perspective from the national scene

to a global arena. The perspective here concerns the view of the world outside Sweden. Three

word clusters are in focus here. The word clusters of "foreign" (utländsk) and "abroad"

(utomlands) are decreasing over the selected years. The same trend is noticed for the word

cluster of "international", which decreases heavily between the years of 1991 and 2001 (from

34 to 6 occurrences). Words such as foreign, abroad and even international all indicate that

national demarcations are important. Therefore the decreased usage of these words is

interesting, especially as the borderless words "global" and "the World" became very

predominant over the investigated period. The share of the references to "the World" and the

"global"-cluster among the selected words expands from 36% in 1981 to 49% in 1991 to a

hegemonic 90% in 2001.7

The shift in perspective from a local Swedish context in 1981 to a wider, more global context

in 2001 can be illustrated by the following excerpts from our sample of CEO letters. In 1981

sample many CEOs pointed out a sense of responsibility and concern for the national

development:

7 The word forms in the "global"-cluster are not compound ones, which is the case of the word forms in the

"world"-cluster. A form like world market is evenly distributed over all three selected years while the form

world economy not is found in the 1981 sample. It is however represented with five forms for 1991 and nine for

2001. The concept of world leading as an adjective form is also uniquely found in the 2001 CEO letters with six

occurrences.

17

(7) Volvo has, due to its size, a responsibility for a large part of the Swedish economy. Volvo

accounted for more than 7 per cent of the total investments in Swedish industry in 1981. Of

Volvo's investments more than a half were made in Sweden. (Our translation, Volvo 1981)

(8.) [..] I have to express my concern for the Swedish political and economical development. The

benefit for Sweden of continued growth of export, international industrial collaboration and

technical development based on the free market conditions cannot be overestimated. (Our

translation, Atlas Copco 1981)

In 2001, the CEO of Volvo (and of Skanska) instead declared that their company had reached

a global position. To have a global position means in this setting to have operations in many

places over the world, and not being particularly dependent on the economic development of a

single country:

(9) Volvo has achieved a strong global position that provides the scale and strength for sustainable

profitability and the achievement of our wanted position. (Our italics, original English version,

Volvo 2001)

(10) The Group's financials are strong, which will enable us to maintain our dividend policy. We

can now build further on our strong global position. Everywhere, Skanska is characterized by

highly dedicated, talented employees, with an unwavering commitment to serving clients. (Our

italics, original English version, Skanska 2001)

Analysis

Our paper clearly shows that the emergence of the CSR discourse has occurred

simultaneously as the discourse about the national development has vanished, and at the same

time the companies have become more international and the CEO letters more global in their

content (see Figure 2 below). We will also argue that the rise of the CSR discourse in several

ways is a part of the globalization. The CSR guidelines from the UN or the OECD are

developed in order to be applicable all over the world. The CSR discourse thus takes a global

perspective and this can lead to a decline of more local discourses of responsibility. Before

the late nineties issues about social responsibility in Sweden was actively mediated in the

national context, now these issues have become much more universally defined.

18

Figure 2. Distribution of words related to CSR and the political and societal discourse in

1981, 1991 and 2001

The critical perspective in the paper evoke us to reflect whether these shifts in discourse usage

also are related to a power shift, in this case away from actors linked to the national arena

such as the State, the Government and national unions. The CSR philosophy emerges in this

setting as an attractive solution for companies who want to "liberate" themselves from

national jurisdiction as in many ways, we argue, the adoption of the CSR philosophy implies

a rejection of social responsibility discourses that are nationally defined. According to the

CSR philosophy it is no longer sufficient to be responsible in relation to a single country, but

it is perhaps neither as necessary as before, due to the fact that social responsibility is more

and more seen as something that companies should do on voluntary basis (self-regulation) in

order to attract customers, employees and investors. Perhaps it is possible to combine social

responsibility with good profits in many cases, or even better, in some cases the image of

being a social responsible company may enhance the profitability. But what about cases when

superior profits and social responsibility do not go hand in hand? In such cases the CSR-

philosophy is not a powerful tool for correcting when insufficient social responsibility occurs.

The possibilities of creating a mass movement of angry consumers are limited, especially over

time. Cases like the boycotting of Shell Petroleum, which made this company to change its

plan of dumping the Brent Spar oil platform in the North Sea (see Grolin 1997) are relatively

rare and thus, they do not represent a powerful tool for correcting bad business behaviour

0

10

20

3040

50

60

70

1981 1991 2001

Sample years

Word

occurr

ences

"CSR" words

"political" words

19

other than occasionally. The CSR-discourse appears instead to provide legitimacy for those

companies who want to behave a bit better than the most ruthless of the actors.

We also want to question whether the CSR have become an important discourse self-

regulating the overall business behaviour, and in particular not in Sweden. Even if the rhetoric

is on rise, we claim that it has not been developed into an important movement, in which

managers on various levels incorporate CSR in their business decision and launch special

activities in the area. More than two years after the invitation from the Government to join the

Swedish Partnership for Global Responsibility, in total only 15 companies have joined, many

of them state owned (see Appendix C). Only two of the members belong to the industrial

sector. H&M, the successful retail company in clothing, is the only company in this list,

which also is included in our sample consisting of the most important companies on the

Stockholm Exchange.

A somewhat higher proportion of the Swedish companies belong to the UN Global Compact

Network. Of the almost 1.500 corporate members, 20 are from Sweden.8 But only a few of

these 20 members provide some information about what they have done in order to meet the

expectations of the guidelines. H&M appears to be the Swedish company that most seriously

have adapted to the CSR philosophy; they presented three different projects in their review of

the CSR-work (one project was an educational centre for former child workers in

Bangladesh). But one committed company, carefully watched by many ethically sensitive

consumers (among them many young women) does not constitute a social movement. On the

whole there exists a disturbing gap between the proud declarations of being a socially

responsible company and to actually provide substantial contributions to the societal

development around the Globe. The engagement in ethical issues is often also accompanied

by disengagement in political issues, as shown in our textual analysis. In Sweden and perhaps

also around the Globe.

8 Dagens Industri, June 1, 2004

20

References

Fairclough, Norman (1992). Discourse and Social Change. Oxford: Blackwell.

Fairclough, Norman (1995). Critical Discourse Analysis. London: Longman.

Grimshaw, Allen D. (2003). Genres, Registers, and Contexts of Discourse. In Arthur C.

Graesser, et al (ed.). Handbook of Discourse Processes. Mahwah: Lawrence Erlbaum

Associates. Pp. 25–82.

Grolin, Jesper (1997) Corporate legimacy and risk society. Lessons from the Brent Spar

conflict. Paper presented at the 14th Nordic Conference on Business Studies, 14-17

August 1997.

Hyland, Ken (1998). 'Exploring Corporate Rhetoric: Metadiscourse in the CEO's Letter.' The

Journal of Business Communication 35(2): 224-245.

Piper, Alison. (2000). 'Some have credit cards and others have giro cheques: 'individuals' and

'people' as lifelong learners in late modernity'. Discourse & Society 11(3): 515-542.

Skulstad, Aud Solbjørg (2002). Established and Emerging Business Genres. Studia

Humanitatis Bergensia 15. Kristiansand: Høyskoleforlaget.

Stubbs, Michael (1996). Text and Corpus Analysis. Oxford: Blackwell Publishers.

Swales, J. M. (1990). Genre analysis : English in academic and research settings.

Cambridge, Cambridge Univ. Press.

Tengblad, Stefan (2003). Expectations of Alignment. Examining the Link between Financial

Markets and Managerial Work. Organization Studies, 25(4): 583-606.

Thomas, Jane (1997). 'Discourse in the Marketplace: The Making of Meaning in Annual

Reports'. The Journal of Business Communication, vol. 34 (1): 47-66.

Yuthas, Kristi, Rodney Rogers and Jesse F. Dillard (2002). 'Communicative Action and

Corporate Annual Reports'. Journal of Business Ethics 41: 141-157.

21

Appendix A. Description of the selected companies

Company Description Founded, year

Atlas Copco Manufacturing/machinery 1873

Electrolux Manufacturing/electric appliances 1919

Ericsson Manufacturing/telecom 1876

H&M (Hennes & Mauritz) Clothing retail 1947

Handelsbanken Bank/finance 1871

Industrivärden Investment company 1944

Investor Investment company 1916

Sandvik Manufacturing/tools 1862

SCA (Svenska Cellulosa Aktiebolaget) Paper producer 1929

SEB (Skandinaviska Enskilda Banken) Bank/finance 1856

Skandia Insurance/finance 1855

Skanska Construction company 1887

SKF (Svenska Kullagerfabriken) Manufacturing/machinery 1907

Trelleborg Manufacturing/rubber and plastics 1905

Volvo Manufacturing/cars, trucks etc. 1926

22

Appendix B. Size of corpora and CEO letters

Number of words in CEO letters Company

1981 1991 2001

Atlas Copco 594 1.109 1.165

Electrolux 1.127 1.139 2.402

Ericsson 906 497 1.612

H&M no CEO letter 613 1.160

Handelsbanken 992 2.034 2.738

Industrivärden 1.115 856 1.353

Investor no CEO letter 1.302 1.389

Sandvik 602 1.037 1.031

SCA 990 926 1.242

SEB no CEO letter 835 1.241

Skandia 1.817 1.446 1.235

Skanska no CEO letter 1.274 847

SKF 2.150 395 746

Trelleborg 726 1.160 2.027

Volvo 1.172 1.051 648

Word total 12.191 15.674 20.836

23

Appendix C. Members of the Swedish Partnership for Global Responsibility (June 2004)

Company Description

ICA A leading private chain of commodities stores

Löfbergs Lila A private coffee brand

The Body Shop An international alternative cosmetics chain

Folksam A cooperative insurance company

H&M (Hennes &

Mauritz)

A large private retailer in clothing

OMHEX A provider of financial exchanges

ITT Flygt A manufacturer of industrial pumps, etc., belonging to the ITT group

Vattenfall A state-owned producer of electricity

KPA An insurance company linked to the municipality sector

SWECO A private company in the building engineering sector

Banco An investor company belonging to the ABM AMRO Bank and specialized on ethical

investments.

V & S A state-owned producer and distributor of liquor (Absolut Vodka, etc)

Lernia A state-owned educational company

Apoteket A state-owned national retail chain for pharmacies

Akademiska hus A state-owned real estate company